Generally, an automatic repeat request (ARQ) is a response message notified by a receiving side to a transmitting side after receiving the data transmitted from the transmitting side. The ARQ informs the transmitting side whether the data was correctly received. Furthermore, the ARQ can be classified into three systems, as shown in FIGS. 1A to 1C, respectively.
FIG. 1A shows a ‘stop-and-wait’ ARQ system, in which a transmitting side waits after data transmission to receive an ACK or NACK message. The transmitting side then sends new data or retransmits former data.
FIG. 1B shows a ‘go-back-N’ ARQ system, in which a transmitting side continuously transmits data regardless of a response from a receiving side. After receiving a NACK signal, the transmitting side retransmits data from a corresponding portion.
FIG. 1C shows a ‘selective-repeat’ ARQ system, in which a transmitting side continuously transmits data regardless of a response from a receiving side. After receiving a NACK signal, the transmitting side retransmits the data corresponding to the received NACK signal only.
Hybrid ARQ (HARQ) is proposed to solve a problem occurring when a larger error occurs over a channel as a higher coding rate (Rc=5/6, 3/4), a high-order modulation (Mod=16-QAM, 64-QAM) and the like are selected due to a demand for a data rate over 2 Mbps, 10 Mbps or higher in a packet transmission communication system.
The erroneous data in transmission is stored in a buffer to have forward error correction (FEC) applied thereto by being combined with retransmitted information in the HARQ system. In contrast, the erroneous data in transmission is discarded in the ARQ system. The HARQ system is a type of system generated from combining FEC and ARQ together. Moreover, the HARQ can be mainly classified into the following four systems.
In the first system, a Type I HARQ system shown in FIG. 2, data is always attached to an error detection code to preferentially detect FEC (forward error correction). If there still remains an error in a packet, retransmission is requested. An erroneous packet is discarded and a retransmitted packet is used with a same FEC code.
In the second system, a Type II HARQ system called IR ARQ (incremental redundancy ARQ) shown in FIG. 3, an erroneous packet is not discarded but is stored in a buffer to be combined with retransmitted redundancy bits. In retransmission, parity bits except data bits are retransmitted only. The retransmitted parity bits are changed each retransmission.
In the third system, a Type III HARQ system shown in FIG. 4, which is a special case of the Type II HARQ system, each packet is self-decodable. The packet is configured with an erroneous part and data to be retransmitted. This system is more accurately decodable than the Type II HARQ system but is disadvantageous in the aspect of coding gain.
In the fourth system, a ‘Type I with soft combining’ HARQ system shown in FIG. 5, a function of data initially received and stored by a transmitting side with retransmitted data is added to the Type I HARQ system. The ‘Type I with soft combining’ HARQ system is called a metric combining or a chase combining system. This system is advantageous in the aspect of signal to interference plus noise ratio (SINR) and always uses the same parity bits of the retransmitted data.
Recently, many efforts have been made to research and develop OFDM (orthogonal frequency division multiplexing) or OFDMA (orthogonal frequency division multiplexing access) suitable for high-speed data transmission over a wired/wireless channel. In OFDM, frequency use efficiency is raised using a plurality of carrier waves having mutual orthogonality. A process of modulating/demodulating a plurality of the carrier waves in a transmission/reception has the same result as performing IDFT (inverse discrete Fourier transform)/DFT (discrete Fourier transform) and can be implemented at a high speed using IFFT (inverse fast Fourier transform)/FFT (fast Fourier transform).
A principle of the OFDM is to reduce relative dispersion in a time domain by multi-path delay spread in a manner of increasing a symbol duration by dividing a high-speed data stream into a plurality of low-speed data streams and by simultaneously transmitting a plurality of the low-speed data streams using a plurality of subcarriers. And, a transmission of data by the OFDM uses a transmission symbol as a unit.
Since the modulation/demodulation in the OFDM can be collectively handled for all subcarriers using DFT (discrete Fourier transform), it is unnecessary to design a modulator/demodulator for each of the individual subcarriers.
FIG. 6 illustrates a configuration of an orthogonal frequency division multiplexing (OFDM) modulator/demodulator. Referring to FIG. 6, a serially inputted data stream is transformed into parallel data streams amounting to the number of subcarriers. Inverse discrete Fourier transform (IDFT) is carried out on each of the parallel data streams. For fast data processing, IFFT (inverse fast Fourier transform) is used. The inverse-Fourier-transformed data is then converted to serial data again to be transmitted through frequency conversion. A receiving side receives the corresponding signal to demodulate through a reverse process.
In a mobile communication system, resources include frequency channels, i.e., frequency bands. Multiple access is a methodology of allocating the limited frequency bands to users for efficient use. Duplexing is a connection methodology of identifying an uplink (UL) connection and a downlink (DL) connection in bi-directional communication. Radio multiple access and multiplexing systems are the basic platform technology of the radio transmission to use the limited frequency resource efficiently and depend on an assigned frequency band, the number of users, a data rate, mobility, a cell structure, a radio environment, etc.
OFDM (orthogonal frequency division multiplexing), which is a sort of MCM (multicarrier transmission/modulation) system that uses several carriers, is a system that parallels input data as many as the number of used carriers to transmit the data loaded on the corresponding carriers. The OFDM is a strong candidate for a radio transmission technology meeting the requirements of a fourth generation mobile communication infrastructure and can be classified into OFDM frequency division multiple access (OFDM-FDMA), OFDM time division multiple access (OFDM-TDMA) and OFDM code division multiple access (OFDM-CDMA) according to a user's multiple access system. Each of the OFDM-FDMA, OFDM-TDMA and OFDM-CDMA systems has its merits and demerits. Moreover, schemes exist to compensate for the demerits.
The OFDM-FDMA (OFDMA), which is suitable for a fourth generation macro/micro cellular infrastructure, has no intra-cell interference, a high efficiency of frequency reuse and excellent adaptive modulation and granularity. Using dispersed frequency hopping, multiple antennas, powerful encoding and the like to compensate for the demerits of the OFDM-FDMA, diversity can be raised and the influence of inter-cell interference can be reduced. The OFDMA can efficiently distribute resources by allocating the number of subcarriers differently according to a data rate requested by each user. Furthermore, the OFDMA can raise the transmission efficiency since it is unnecessary for each user to perform initialization using a preamble prior to data reception like OFDM-TDMA. In particular, the OFDMA, which is suitable for a case using numerous subcarriers (e.g., a case wherein an FFT size is large), is efficiently applied to a radio communication system having a relatively wide cell area. Also, the frequency-hopping OFDMA system is used in raising a frequency diversity effect and obtaining an intermediate interference effect by overcoming a case where a subcarrier in deep fading exists in a radio channel or a case where there exists subcarrier interference caused by another user. FIG. 6 shows the OFDMA system, in which an allocated grid performs frequency-hopping in a frequency domain according to a time slot.
FIG. 7 is a structural diagram of a data frame in an OFDMA radio communication system according to the related art. Referring to FIG. 7, a horizontal axis is a time axis represented by a symbol unit and a vertical axis is a frequency axis represented by a subchannel unit. The subchannel refers to a bundle of a plurality of subcarriers. In particular, in an OFDMA physical layer, active carriers are divided into groups to be transmitted to different receiving ends, respectively. Thus, the group of subcarriers transmitted to one receiving end is called a subchannel. In this case, the carriers configuring the subchannel can be adjacent to each other or can be spaced uniformly apart from each other.
A slot allocated to each user, as shown in FIG. 7, is defined by a data region of a two-dimensional space, which is a set of consecutive subchannels allocated by a burst. In the OFDMA, one data region, as shown in FIG. 7, can be represented as a rectangle determined by time and subchannel coordinates. Such a data region can be allocated to a specific user's uplink. Also, a base station can transmit such a data region to a specific user in downlink.
In the related art OFDM/OFDMA radio communication system, in case that data exists to be transmitted to a mobile subscriber station (MSS), a base station (BS) allocates a data region to be transmitted via a DL-MAP (downlink-MAP). The mobile subscriber station receives the data via the allocated region (DL bursts #1 to #5 in FIG. 7).
In FIG. 7, a downlink subframe starts with a preamble used for synchronization and equalization in a physical layer and a structure of an entire frame is defined via broadcast-formatted downlink MAP (DL-MAP) and uplink-MAP (UL-MAP) messages defining locations and usages of bursts allocated to the uplink and downlink, respectively.
The DL-MAP message defines the usage allocated per burst to a downlink interval in a burst-mode physical layer, and the UL-MAP message defines the usage of the burst allocated to an uplink interval. In an information element (IE) configuring the DL-MAP message, a downlink traffic interval is identified on a user end by DIUC (downlink interval usage code) and position information (e.g., subchannel offset, symbol offset, subchannel number, symbol number) of the burst. Meanwhile, in an information element configuring the UL-MAP message, the usage is determined by UIUC (uplink interval usage code) per CID (connection ID) and a position of a corresponding interval is regulated by ‘duration’. In this case, the usage per interval is determined according to a value of the UIUC used in the UL-MAP. Each interval starts from a point having a distance away from a previous IE start point, wherein the distance is as far as the ‘duration’ regulated by the UL-MAP IE.
A DCD (downlink channel descriptor) message and a UCD (uplink channel descriptor) message include modulation types, FEC code types and the like as physical layer associated parameters to be applied to the burst intervals allocated to the downlink and the uplink, respectively. Also, necessary parameters (e.g., K, R, etc. of R-S code) according to various forward error correction code types are provided. These parameters are given by burst profiles provided for the UIUC (uplink interval usage code) and DIUC (downlink interval usage code) in the UCD and DCD, respectively.
In the OFDMA communication system, the burst allocating method can be classified into a general MAP method and a HARQ method according to whether the HARQ system is supported.
The burst allocating method of the general MAP in downlink teaches a rectangular shape, as shown in FIG. 7, configured with time and frequency axes. Namely, it teaches a start symbol number (symbol offset), a start subchannel number (subchannel offset), the number of used symbols (No. OFDMA symbols) and the number of used subchannels (No. Subchannels). Since a method of allocating bursts to a symbol axis sequentially is used in the uplink, uplink bursts can be allocated by teaching the number of the used symbols only.
FIG. 8 is a diagram of a data frame according to a HARQ MAP. Referring to FIG. 8, in the HARQ MAP, a method of allocating bursts along a subchannel (subcarrier) axis sequentially is used in both uplink and downlink, which is different from that of a general MAP. In the HARQ MAP, a length of a burst is informed only. In this method, bursts, as shown in FIG. 8, are sequentially allocated. A start position of a burst corresponds to a position where a previous burst ends and occupies a radio resource amounting to an allocated length from the start position. The method explained in the following relates to a method for allocating bursts in an accumulative form along a frequency axis. A method for allocating bursts along a time axis follows the same principle.
In the HARQ MAP, a MAP message may be divided into a plurality of MAP messages (e.g., HARQ MAP#1, HARQ MAP#2, . . . , HARQ MAP#N) so that each of the divided MAP messages can have information of a random burst. For instance, a MAP message #1 can include information of a burst #1, a MAP message #2 can include information of a burst #2, and a MAP message #3 can include information of bursts #3˜#5.
As mentioned in the foregoing description, the OFDMA system uses the HARQ MAP to support the HARQ. Since a HARQ MAP pointer IE is included in the DL MAP, a method exists for allocating bursts sequentially along a downlink subchannel axis in the HARQ MAP if a position of the HARQ MAP is informed. A start position of a burst corresponds to a position where a previous burst ends and occupies a radio resource amounting to an allocated length from the start position, which is applied to the uplink as it is.
In the HARQ MAP, control information should be informed. Table 1 shows a data format of a HARQ control IE to indicate the control information.
TABLE 1SyntaxSize (bits)NotesHARQ_Control_IE( ){Prefix10 = temporary disable HARQ1 = enable HARQIf(Prefix == 1){AI_SN1HARQ ID Seq. NoSPID2Subpacket IDACID4HARQ CH ID} else{reserved3Shall be set to zero}}
The control information includes AI_SN, SPID, SCID, etc. The AI_SN is a value, which is toggled between ‘0’ and ‘1’ if a burst transmission is successful over a same ARQ channel, for indicating whether a transmitted burst is a new burst or corresponds to a retransmission of a previous burst. Four kinds of redundancy bits are reserved for the data bits put in each burst for the HARQ transmission. The SPID is a value for selecting a different redundancy bit during each retransmission. The SCID is a HARQ channel ID.
An ACK signal region of the uplink is informed via an ACK/NACK signal whether the transmitted data burst was successfully received. If a mobile subscriber station receives a burst at an ith frame, the ACK/NACK signal is sent to the ACK signal region of the uplink of an (i+j)th frame. A value of ‘j’ is sent by the UCD. In allocating the ACK signal region, one method exists for allocating the ACK signal region to the uplink for each HARQ message. Another method exists wherein at least two of a plurality of HARQ MAP messages of a frame uses one ACK signal region.
A method wherein slots of an ACK/NACK signal of a burst indicated by a HARQ MAP message is sequentially informed by deciding an HARQ ACK region of a frame as one is explained in detail as follows.
FIG. 9 is a diagram of a method for allocating an HARQ signal region in an HARQ MAP message. In an HARQ MAP message, an ACK signal region is allocated to an uplink using a start position of the ACK signal region and four kinds of information (OFDMA symbol offset, Subchannel offset, No. OFDMA Symbols, No. Subchannels). Each mobile subscriber station sequentially inputs an ACK/NACK signal to the ACK signal region (FIG. 9) allocated to the uplink for indicating whether a respective burst has been successfully received. A start position of the ACK/NACK signal corresponds to a position next to that of the previously received ACK/NACK information. A sequence of ACK/NACK signals follows a burst sequence of a downlink within the HARQ MAP message. Namely, like the sequence of bursts #1 to #7, the ACK/NACK signals within the allocated HARQ ACK region of the uplink are sent in a sequence that corresponds to the sequence of the bursts #1 to #7.
Referring to FIG. 9, a MAP message #1 includes allocation information of bursts #1 and #2, a MAP message #2 includes allocation information of bursts #3 and #4, and a MAP message #3 includes allocation information of bursts #5 to #7. Mobile subscriber station #1 (MSS#1) reads the information of the burst #1 in the contents of the MAP message #1 and then informs an initial slot within the HARQ ACK signal region indicated by an HARQ MAP message whether the transmitted data was successfully received. MSS#2 knows its position within the HARQ ACK signal region by recognizing that it is sequentially next to that of the ACK/NACK signal slot of the burst #1 within the ACK signal region (position within the HARQ ACK region is known by incrementing a count of the burst #1 within the contents of the MAP message #1). MSS#3 knows its position within the HARQ ACK region by calculating a total amount of slots of the bursts #1 and #2 of the MAP message #1. Thus, the positions within the HARQ ACK region can be sequentially known.
In case that one mobile subscriber station supporting a multi-antenna to an area of the downlink burst loads data on the same area to transmit or in case that several mobile subscriber stations load data on the same area to transmit, the ACK signal is sent only if there is no error in a cyclic redundancy check (CRC) for all layers. Otherwise, the NACK signal is sent. In this case, a layer means a coding unit of the transmitted data and the number of layers directly corresponds to the number of antennas depending on how the data is transmitted. For example, if the entire data to be transmitted is coded. A CRC is then inserted in the coded data. This is then divided by the number of antennas. If the divided data are transmitted via all the antennas, the number of layers is equal to one. In another example, if data to be loaded on each antenna is coded. A CRC is then inserted in the coded data. If the coded data is transmitted, the number of layers is equal to the number of antennas (cf. FIG. 10). The above-explained situation is applicable to a case where a mobile subscriber station transmits a burst in uplink and a case where a base station having received the burst sends an ACK signal in downlink.
The above-explained related art method can be simply applied to a system that is not a multi-antenna system. Yet, in case of the multi-antenna system, the related art method brings about a waste of resources. For example, if a base station detects a case that two mobile subscriber stations #1 and #2 load their data on the burst #2, the number of layers is 2. Furthermore, the burst of the mobile subscriber station #1 is not erroneous but the burst of the mobile subscriber station #2 is erroneous. The base station then sends a NACK signal to both the mobile subscriber stations #1 and #2 according to the aforesaid principle of the related art. If so, both of the mobile subscriber stations should send the data again. Consequently, the errorless data of the mobile subscriber station #1 is discarded to be retransmitted, which is a waste of resources. Moreover, the same problem of the uplink can be directly applied to the case of the downlink.